NVIDIA Tesla K10
Optimised for single precision applications, the Tesla K10 is a throughput monster based on the ultra-efficient GK104 Kepler GPU.
| Specification |
| GPU Nodes |
| 2 Kepler GK104s |
|
| GPU Computing Applications |
| Seismic processing, signal and image processing, video analytics |
|
| Memory size (GDDR5) |
| 8 GB (4 GB per GPU) |
|
| CUDA cores |
| 3072 (1536 per GPU) |
|
| Memory bandwidth (ECC off) |
| 320 GB/sec (160 GB/sec per GPU) |
|
| Peak double precision floating point performance |
| 190 Gigaflops (95 Gflops per GPU) |
|
|
| Peak single precision floating point performance |
| 4577 Gigaflops (2288 Gflops per GPU) |
|
|